AI agent skills for academic research workflow
An installable AI-agent skill pack (Claude Code / Codex / OpenCode) that drives a research, competition, or innovation project from vague idea through literature review, experiments, LaTeX writing, figures, and patent materials.
Same problem, same approach. Swapping one for another is a config change, not a rewrite.
Markdown-only, no-framework peer; pick it when you want cross-model review loops and unattended overnight ML runs instead of Light-skills' human-confirmation gates.
ARIS ⚔️ (Auto-Research-In-Sleep) — Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment automation. No framework, no lock-in — works with Claude Code, Codex, OpenClaw, or any LLM agent.
Chinese-language drop-in peer for advisor-style idea refinement and submission coaching; check its non-standard license before forking.
将博导十年科研经验炼化为可直接调用的 AI 技能。从 Idea 构思到论文投稿，你的 AI 科研副导师。
Pick it when you also want the coding and experiment half of the loop and multi-CLI support, versus the seed's writing- and figure-centric skills.
Semi-automated research assistant for academic research and software development. Supports Claude Code, Codex CLI, Kimi Code CLI, and OpenCode across ideation, coding, experiments, writing, and publication.
Choose it when you need MATLAB scientific computing and Word/PPT deliverables bundled in; it is thinner on patents and competition demos.
本仓库包含三个面向学术科研人员的Skills，覆盖从文献阅读、论文写作到科学计算的完整研究工作流。office-academic-skill 负责论文阅读报告与学术 PPT/Word 文档生成；research-writing-skill 提供论文写作、润色与审稿回复辅助；scientific-toolkit-skill 整合 MATLAB/Python 科学计算与期刊级图表生成。三个 Skill 专为中文科研用户设计，可在 Claude Code 和 codex 中一键调用。
Same-install-model peer with 182 skills; choose it when your work needs bioinformatics or drug-discovery skills beyond the seed's scope.
A curated, multilingual library of 182 installable AI agent skills for end-to-end academic research—spanning literature discovery, scientific writing, grant development, bioinformatics, drug discovery, clinical research, machine learning, and data analysis.
Codex-native peer covering research-to-finalize; pick it if your stack is Codex-first and you want explicit human-in-the-loop checkpoints.
Codex-native Academic Research Skills suite for human-in-the-loop academic research workflows
The most widely adopted Claude Code peer for the same pipeline; choose it for a polished four-stage flow, but verify its non-standard license.
Academic Research Skills for Claude Code: research → write → review → revise → finalize
Model-agnostic skill library; pick it when you want ML-engineering and research skills beyond the seed's academic-writing and figure focus.
Comprehensive open-source library of AI research and engineering skills for any AI model. Package the skills and your claude code/codex/gemini agent will be an AI research agent with full horsepower. Maintained by Orchestra Research.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
A runnable Python framework that executes the whole idea-to-paper loop unattended; choose it over composable skill files if you can live with 9-month-old code and a non-standard license.
The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery 🧑🔬
Hands-off idea-to-paper autonomy with no per-step checkpoints; take it when you want a pipeline to run, not prompts that ask you questions.
Fully autonomous & self-evolving research from idea to paper. Chat an Idea. Get a Paper. 🦞
Agentic tree-search variant of the same autonomous pipeline; a heavyweight Python system you execute rather than skills you install.
The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search
Self-contained end-to-end research pipeline; only worth swapping if you accept ~13-month-stale code and want a framework over composable skills.
Agent Laboratory is an end-to-end autonomous research workflow meant to assist you as the human researcher toward implementing your research ideas
Blueprint-driven runtime that orchestrates the pipeline; take it when you need a scheduler for long autonomous runs rather than interactive skills. No license declared.
A blueprint-driven AutoResearch runtime for orchestrating AI research workflows from idea generation and experiments to paper writing and peer review.
Autonomous research assistant runtime; pick it when you want an agent that runs the study itself rather than guided skill steps.
🦞+🔬 NanoResearch: The Autonomous AI Research Assistant
Self-organizing agent-team runtime for long-running experimentation; use it for continuous autonomous runs, not interactive authoring help.
AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation
Claude-Code-driven reimplementation of the Kosmos paper for autonomous discovery; paper-specific and six months stale.
Kosmos: An AI Scientist for Autonomous Discovery - An implementation and adaptation to be driven by Claude Code or API - Based on the Kosmos AI Paper - https://arxiv.org/abs/2511.02824
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
A LaTeX/Beamer + R Claude Code template for econ-style papers; narrower and more personal than the seed's full workflow.
A ready-to-fork Claude Code template for academics using LaTeX/Beamer + R. Multi-agent review, quality gates, adversarial QA, and replication protocols.
Pre-submission adversarial review only; add it when the missing piece is a reviewer verdict, not general research workflow.
Pre-submission AI review stress-test for research papers. A Claude Code skill: review, verdict, revise, verify.
A single-principle prompt pack for prose style; far narrower than the seed, useful only for stripping defensive hedging.
极致轻量｜阻止论文的防御性写作 | Anti-defensive writing for academic papers — 学术写作原则 Skill + prompt pack (CN/EN)
Planning-and-composition skill pair only; narrower than the seed and last touched 9 months ago.
Systematic framework for planning and writing academic papers using Claude Code. Includes strategist (planning) and composer (writing) skills with quality checkpoints.
Journal-targeting skill packs; use it when venue formatting and submission checklists are the gap, not the whole research workflow.
Journal-specific Claude Code/Codex skill packs covering mainstream journals — AER, QJE, Nature, Cell, 管理世界, 经济研究 & 200+ more — your fast track to getting published. ｜ 覆盖主流期刊的 Claude Code/Codex 期刊技能包，从选题、识别策略到表格规范与审稿回复全流程，助你快速发论文。
Does the same job, but ships as an app. Useful to a person, not swappable into a codebase.
An offline-first desktop writing workspace bundling LaTeX, Python and skills; adopt it as an app, not as skills dropped into your repo.
An offline-first scientific writing workspace powered by Claude. LaTeX + Python + 100+ scientific skills all running locally.
These projects were analysed and named Light-skills among their alternatives. The relationship is not symmetric — how Light-skills rates them is a separate judgement, made when Light-skills is analysed in its own right.
calls Light-skills “choose for a general research/competition/innovation skill pack that also covers academic work”
Academic Research Skills for Claude Code: research → write → review → revise → finalize